Marine Recreational Information Program (MRIP) - Past, Present
and Future
Around 2003, as recreational fishing regulations became more restrictive, many recreational fishermen
began to complain about the Marine Recreational Fishery Statistic Survey (MRFSS), the current way NOAA
captures information about anglers’ fishing activity. The Recreational Fishing Alliance and the Watermen
Association of NY and NJ as well as many other groups complained about the inability of the MRFSS to
provide catch and effort data of the quality needed for fishery management allocations. I became involved
with the Atlantic Coastal Cooperative Statistics Program (ACCSP) as an ACCSP Advisor in January 2004 in
the hope to understand the MRFSS and the fishery in general.
Then NOAA Fisheries Chief, Dr. Bill Hogarth, that same year commissioned the National Academy of
Science’s National Research Council (NRC) to gain unbiased scientific advice that the agency could use to
improve its recreational fishery data collection program. The report was done and released in April 2006.
The constituents that were part of the review included the American Sportfishing Association, Coastal
Conservation Association, Recreational Fishing Alliance, National Association Charterboat Operators,
Golden Gate Fisherman Association, Jersey Coast Angler Association and Fishing Rights Alliance among
others.
The reauthorization of the Magnuson-Stevens Act was presented in Congress in November 2005; the
original bill was S2012 and was passed in the senate June 2006. The House bill H.R. 5946 was presented
July of 2006 and passed the House September 2006 and the Senate December 2006. One recommendation
of the NRC study that anglers may be most familiar with was for the creation of a recreational registry to
improve the collection of recreational data that would replace the random digit dialing (RDD) method of
determining fishing effort.
On August 7 – 9, 2007, about 90 recreational fisheries experts gathered in St. Pete Beach, Florida to kickoff
an effort to improve marine recreational fishing information. The overall goal of the initiative is to develop
a better recreational survey to satisfy the needs of assessment scientists and fishery managers, while at the
same time be accepted and transparent to the recreational fishing community. This next generation
program is called the Marine Recreational Information Program (MRIP) and it will eventually replace the
existing MRFSS.
A wide variety of expertise and interests within the recreational fishing community were represented at
the workshop including marine biologists, statisticians, stock assessment scientists, fisheries managers,
and constituent representatives who represented all sectors of the recreational fishery. The Rhode Island
Saltwater Angler Association (RISAA) sponsored me to attend that meeting Participants came from the
U.S. Virgin Islands, American Samoa and Puerto Rico, as well as Hawaii, Alaska and nearly every coastal
state. Individuals from these sectors are helping plan the initiative. Right now, there are five working
groups - Analysis, Design, Highly Migratory Species, For-Hire, and Data Management and Standards and
three teams – Operations, National Saltwater Registry, and Communications and Education.
One important piece of the initiative is an angler registry or saltwater fishing license. The MagnusonStevens Fishery Conservation and Management Act required NOAA to create a national angler registry,
similar to a phonebook of saltwater anglers. The exception being that states with a suitable saltwater
fishing licenses, registrations or permits can exempt their anglers. The requirement for individual anglers
to register went into place on January 1, 2010. As of today, all states from Maine to New York either have a
license or registry program in effect meaning New England anglers do not need to register with NOAA so
long as they are in compliance with state requirements.
We need to step back a minute to point out the limitation of MRFSS. In the late 1970’s the MRFSS was
designed to help with assessments and coastwide mortality. The MRFSS project design was the summing of
random strata allowing for the estimator converging to a true estimate at the coastwide level. Somewhere
along the line Management went from assessment to using the data for quota management which was not
the original design of the MRFSS. This was also the MRFSS Achilles heel because when you post-strata data,
the MRFSS then begins to unravel. I would like to express concepts with analogies. The fishery manager
first wanted good gas mileage car and NOAA design a Honda Civic. Later the fishery managers need to tow a
28 foot boat. Since the Honda Civic can’t tow the boat does that make the Honda Civic fatally flawed, NO.? It
just makes the Honda Civic the wrong tool for the task at hand. Now NOAA is now building a truck (MRIP)
to do the job which to say is designing a tool that can be used to do quota management.
Recreational fishermen have trouble with the concept of surveys and are always trying to compare poststrata information of the MRFSS with census data; it is hard to look at post-strata data and use it as a point
source rather than a variant with a deviation. The analogy I like to use is that the MRFSS is like a picture in
the form of Jpeg. The coastwide picture of the MRFSS would look like a green field with a big tree in the
middle. As I enlarge the picture to get a closer look at detail which is like someone looking at post-strata
data of the MRFSS, the picture then becomes out of focus and as I continue to magnify the picture gets
distorted until I only see out of focus dots. The analogy with census data the picture in a form of adobe
acrobat format is a green field with a tree but sections of the picture is missing (since summing a census
you always miss pieces of data) but if you enlarge the pieces of the picture that are not missing you can see
the individual pixels which are not distorted. The flexibility of MRIP should result in a methodology to
combine both methods, which can represent one picture with depth of data to give the information needed
for Quota Management and Annual Catch limits and Accountability Measures.
So what has happen in the last four and half years? MRIP had an original budget funding at 3.5 million in
FY 2008 and 6.2 million in FY 2009. In both FY 2010 and FY 2011 the MRIP program received 9 million.
The Strategic Approach was to move forward with the NRC recommendations provided in the report. What
are the strategic approaches? Evaluate possible biases in the current MRFSS surveys and assess design
changes that could eliminate those biases. In other words, the NRC study identified a series of untested
assumptions in the surveys the Agency was using. NOAA is working to correct faulty assumptions and
produce more accurate numbers. They are going about this by conducting pilot studies to test new
sampling and estimation methods and confirm their effectiveness. They will then implement those
improved methods more broadly once they are evaluated and prove to be better than existing methods.
Where are we now with the MRIP program? There are four improved survey designs with data quality
currently being work on by the MRIP team. One improvement is developing new estimation methods for
current survey and second is testing more efficient telephone and mail survey designs that utilized angler
registries for sampling. Third is to improved sampling design for shoreside survey of angler catches and
last a pilot study to test mandatory electronic logbook reporting for the Gulf of Mexico for-hire fishery that
can be validated with shoreside sampling.
The chief concern about the MRFSS by the NRC report was for potential bias. Potential for bias is
the result of unaccounted for factors or untested assumptions that can skew the catch numbers
higher or lower. The NRC noted that “Identifying and eliminating the sources of bias is a
fundamental requirement for the provision of reliable estimates.” The NRC criticism was that
the catch estimation methods don’t account for complex design of the shoreside survey. Another
criticism by NRC was shoreside survey sampling design allowed flexibility to increase
productivity, but ignored possible impacts on estimation which increased complexity and risk of
sampler errors. Among the NRC’s recommendations was to eliminate potential bias by properly
weighting the data collected from anglers and eliminating some flexibility in the shoreside
sampling design.
What’s changing from MRFSS to MRIP? First it is matching sampling designs by improving both how
information is collected from anglers and how it is used to generate estimates. MRFSS assumed our
shoreside sampling was random and therefore representative of the whole population of anglers, a critical
component of a sound survey. MRFSS assumed that catch sampled during peak times could accurately
estimate catch across and entire 24-hour period. These assumptions introduced bias that could skew catch
estimates. Some of those assumptions included the amount of fish anglers were catching at different
locations and during different times of day. The improved MRIP estimation method corrects those faulty
assumptions and ends up producing more accurate numbers as a result.
Complimenting this new, more statistically sound estimation method, will be new protocols for how
anglers are interviewed at docks, piers, boat ramps, and beaches. . Under these new rules, anglers will now
be sampled throughout all parts of the day, including at night. Sampling during both peak and non-peak
times will enable NOAA to more accurately estimate catch across a whole day.
Another thing that will change is that samplers will have less discretion in where they go and who they
interview. MRFSS samplers’ site assignments were selected based on the pressure rating by mode, day,
type and month. In the past samplers had discretion to sample alternate mode trip (shore, boat, or for-hire)
that they observed occurring at the assigned site. Since selection probability of the site for alternate modes
sampled can’t be determined. Therefore, it is not possible to weight the estimate to account for the true
selection probability. MRIP will collect data for “alternate mode” trips sampled at primary assigned sited
will not be used in new estimates. Now samplers will only collect data for mode for which the site was
selected.
Last MRFSS site assignments were based on Probability Proportional to Size (PPS) sampling based on the
pressure rating by mode, day, type, and month. However, samplers were allowed to move to an alternate
site of their choice after 2 hours if there was little or no activity at the assigned site. Selection probabilities
of the alternate site were not known. Therefore, the estimates were not weighted properly to account for
the alternate site selection probabilities. MRIP selection probabilities of the alternate site have been
estimated (model-based method), and the estimated have been reweighted to account for them. Now,
samplers will conduct interviews at a specific cluster of sited in a specific randomized order for the full
time period ensuring more structured sampling and less sampler discretion.
In late January, 2012, NOAA began using the new catch estimation method. The new method will allow for
the recalculation of past estimates going back to 2004 and will be used to generate catch estimates moving
forward. We will see point estimates changes but it will be randomize some estimates go up, some go down
and some stay about the same. We will see the precision of catch rate estimated increased (PSE) because
this better prediction of variability of the point estimate from the data collected. Regardless of how the
numbers change, the estimates should be more accurate because they were calculated using new
methodology that’s free of bias. And while this new methodology is a big deal, it’s only one step in the
larger process of improving the overall quality of the numbers. For example, NRC also had criticisms of
effort estimates from are offsite sampling methods that rely on telephone interviews which are
complicated by the increasing use of cellular telephones. The existing random digits dialing survey also
suffers from the low proportion of fishing households among the general population and may allow bias in
estimation from its restriction to coastal counties only use. The NRC recommended not relying solely on
these random calls as it assumes that all anglers may have the opportunity to be surveyed. The NRC
recommended, and NOAA is currently undertaking pilot projects that utilize more than one so-called frame
(e.g. the National Saltwater Angler Registry or U.S. postal service directory) as a way to contact anglers
about the number of trips they take. MRIP is testing three survey methods: Phone surveys with angler
license directory only, phones surveys with random digit dialing and license directory and mail surveys
with postal service addresses and license directory.
The revised catch estimates using the new methodology were released this January and will begin being
used by managers this year. Also the NOAA will be evaluating results from a number of pilot projects
including, a new electronic logbook reporting system for charter boats and headboats, an enhanced angler
dockside survey to complement the improved catch estimation methodology, an improved survey utilizing
the National Saltwater Angler Registry which gathers angler trip data, and ways to support more frequent
reporting and posting of estimates.
You may start to see many of these improvements beginning in 2013. In addition, NOAA has committed to
investing in increased sampling to improve precision and the timeliness of the reporting.
This should be the end of this article, but there is another piece of the puzzle, I haven't even talk about
which is the Atlantic Coastal Cooperative Statistics Program (ACCSP). This is the part of the puzzle I’m most
involved in since I’m a Rhode Island recreational Advisor on the ACCSP and also sit on the Recreational
Technical Committee representing the Advisory Committee. The current elected chairman is Rick
Bellavance from Rhode Island representing Commercial and for-hire interest. The ACCSP receives data and
is the data warehouse of recreational, Commercial and biological data and by-catch for the Atlantic coast.
Why is the ACCSP an important part of the MRIP program? Although NOAA provides the design of the old
MRFSS and the new MRIP program, the Pacific and the Gulf Coast has the Rec-Fin and Gulf-Fin respectively
administrate their program regionally to the States on the Atlantic coast NOAA directly administrate the
program to the states. NOAA asked the ACCSP if they would like to be the regional administrators for
Maine down to Florida. In the pass the ACCSP Program Design was coordinated with the MRFSS, The
Recreational Technical Committee has been task for the last 18 months to rewrite the standard so ACCSP
can have minimum standards to administrator the new MRIP program to the Atlantic States. These new
standards will have to be approved by the Operations and Advisory Committee which will be done later
this year and should be available for the Coordination Council for final approval before the last ASMFC
meeting in 2012.
Some of the new standards that the ACCSP has tackled are what states should do and when to do wave one
sampling since currently one North Carolina on the east coast does wave one sampling. The committee is
also addressing when to change from a 2 month wave sampling to monthly sampling. Standard to create instate spit geographically boundaries. State can create in-state splits to separate many specie into multiple
stock are used as a measure to define division that will benefit coastal stock assessment as oppose to
individual states. Examples are Cape Hatteras, Cape Cod and Delaware Bay.
ACCSP developed precision standards on the level of sample needs to increase resolution. The new
standard will support State conduct in collecting catch data in which a state performs the intercept by
department personnel or interns. The for-hire mode have a standards for both survey and census method
of collection data. Last the ACCSP supports innovative approaches in method for collecting data on catch
and effort. These innovative approaches can but not limited to voluntary self-reporting system and as well
as cameras to captures discards since the future could offer new tool to improve data collection.
I thank everyone who was involved in the program either at MRIP or ACCSP. I would like to give specific
thanks to Forbes Darby (NOAA outreach), Beverly Sauls (Florida FWC), Kevin Sullivan(NH Fish and Game
Dept.), Kathy Knowlton (GADNR CRD),Dan Costa (RIDEM)and Geoff White (ACCSP Staff) for their
dedication and hard work to the MRIP program for the Atlantic Coast.
This article is tough to digest but is the complete picture from past to present. Most recreational fishermen
thought that the MRIP would happen quickly but that not going to happen. MRIP is also not a cosmetic
change either. MRIP is a complete fundamental change of collecting and analyzing the data which is so
complex it had to be developed and phased in over time. MRIP is not a static program unlike the MRFSS but
will continually change by developing methodology to use both methods of surveys and census, which can
represent one picture with depth of data. The new MRIP in the next 24 months hopefully will give
information so fishermen can believe, scientists can trust and fishery managers can use.
Michael J Bucko
155 Brackett Ave
Tiverton RI 02721